Dataviz final project

Lucas Artaud & Iswarya Sivasubramaniam DIA 1

Imports

Context and motivation

The data set chosen is a statistical data on factors influencing Life Expectancy. The data comes from the World Health Organization over a 15-year period.
Many studies in the past have explored the factors influencing life expectancy, centering around demographic variables, income composition, and mortality rates. However, these studies often neglected the impact of immunization and the Human Development Index. Additionally, some prior research relied on a one-year dataset for all countries but doing this research for a period of time of 15 year enables us to visualise the changes over time. This data set allows us to do a country based observation to identify the main factors that are contributing to lower life expectancy.
This dataset encompasses health factors for 193 countries from 2000 to 2015, with 22 columns and 2938 rows. Our preliminary analysis indicates that the population, Hepatitis B, and GDP columns contain the majority of missing data. Rather than removing all missing values and losing valuable information, we have opted to retain the incomplete rows.
The project's motivation is to analyze various factors influencing life expectancy, comparing them on different scales such as continents and development statuses.
The objective is to gain insights into factors affecting life expectancy, guiding public health interventions and policies. Targeted healthcare initiatives could be guided, for instance, by identifying particular regions or demographic groups experiencing challenges with life expectancy. Furthermore, knowledge of how social determinants, economic variables, and immunizations affect life expectancy can support evidence-based decision-making at the national and international levels.

Columns

Dataset Analysis

Creating new columns

With the library pycountry_convert we are going to create a new column "Continent" that will correspond to the continent of the country. And with Neonatim we are going to generate the latitude and longitude for each country in order to create maps.

Visualisations

General Analysis

1. Average life expectancy and population over the years

Between 2000 and 2015, we can see that the average life expectancy has increased from 67 to 72 years. For the average population, it is inconsistent over time. In fact, we can observe ups and downs, especially during 2008 and 2010.

2. Life Expectancy over the years of the top 5 and bottom 5 countries

The top 5 countries having the best average on life expectancy over the years are France, Sweden, Iceland, Japan, and Switzerland. The bottom 5 countries having the worst average on life expectancy over the years are Sierra Leone, Malawi, Angola, Central African Republic, and Lesotho. We can notice that the top 5 countries have an increase in life expectancy (from 81-88), but it stabilises from 2009, whereas for the bottom 5, we can see a considerable growth (from 39 to 51 for certain countries). We can also notice that the top 5 countries belong to the northern hemisphere, unlike the bottom 5 that are African countries.

3. Violin Plot for Life Expectancy by Continent

We can conclude that Africa is the continent where life expectancy is low, so the authorities should concentrate on this continent. The second continent having a low average is Asia.

4. Pie chart for the distribution of countries by Status

This dataset contains 17.4% of developed countries and 82.6% of developing countries. This is a notable ratio because, typically, developing countries have lower life expectancy so we can study them in detail.

5. Comparing the GDP from 2000 to 2015 by the status

Over the years, there is a significant gap between the GDP of developed countries and developing countries, taking into account the fact that we have only 17% of developed countries. The gap between GDP and status is considerable. In 2000, we have a gap of 12,842 USD to 20,057 in 2014. There is an increase in GDP on both sides, but the increase is greater for developed countries.

6. Comparing the life expectancy from 2000 to 2015 by the status

We notice an evolution in life expectancy on both sides, but values are more scattered in developing countries compared to the concentrated values in developed countries.

Correlation study

7. Correlation map in order to study the columns that are influencing the life expectancy

From this correlation matrix, we can see that Schooling, Income composition of resources, and BMI are highly correlated to life expectancy. It means that they influence the growth of life expectancy. Let's concentrate on the analysis of these columns.

8. Correlation between schooling and life expectancy by continent

We can see that we have a clear correlation line between schooling and life expectancy. The more the years of schooling are, the better life expectancy is.

9. Correlation between income ressources and life expectancy by continent in 2014

Here we are only concentrating on 2014 because 2015 has many missing values; moreover, it enables us to clearly see the correlation. We can note that most countries in Africa have an Income Composition Resources of 0.34-0.59 and a life expectancy of 48-68. For European countries, we have a higher income composition of resources and a better life expectancy. This explains the correlation between both criteria; the more the income composition of resources, the better is the life expectancy.

Health study

10. Average BMI by continent

TThe body mass index (BMI) is a measure that uses your height and weight to work out if your weight is healthy. Compared to the other continents, Africa has the lowest average BMI score. This can explain the fact that it has a lower life expectancy. Indeed, a lower BMI means that they are unhealthy. This can be caused by malnutrition and may provoke earlier death.

11. Thinness between 1-19 years old accross countries

In order to analyse our hypothesis made in the last visual, we have created this map representing thinness between 1-19 years old across countries. We can see that South Asian countries (like India, Pakistan) have the highest number of thinness between 1-19 years old. Compared to others, African countries also have a relatively high number of thinness between 1-19 years old, but we can also see that this has improved a little over the years. This can be an explanation for the BMI value.

12. Violin plot on Alcohol Consumption by continent

Europe is the continent where alcohol consumption is high compared to other continents. We can also see that it is highly spread.

13. Comparision on the evolution of HIV and Measles

Healthwise, worldwide for this 2 diseases, we can see there is a significant fall in the number of HIV and measles. This decrease can explain the increase in life expectancy worldwide.

14. Map on the evolution of Adult Mortality

Over the years, African countries and Asian countries have the highest number of adult deaths. This can be explained by health development, malnutrition, and also geopolitical situations (we don't have more information about this third point).

15. Map on the evolution of under five death to compare

Over the years, India and China, which are the most populated countries in the world, have a high number of under-5-year-old deaths. We have already seen that India also had the highest number of thinness between 1-19. These two elements can be correlated.

Conclusion

In conclusion, this study enabled us to see the evolution of life expectancy in different parts of the world and, more importantly, to identify the main factors influencing it and the continents/countries facing difficulties. Even though overall life expectancy has increased between 2000 and 2015, some parts of the world are less developed than others in certain aspects.
There is a significant gap between developed and developing countries, especially in GDP and life expectancy. Notably, Africa stands out with the lowest average life expectancy, demanding focused attention. To enhance life expectancy, prioritizing investments in education and income generation is imperative.
On the health aspect, we observe significant progress in the reduction of deaths from HIV and measles over the past 15 years. However, challenges persist in African and Asian countries, marked by lower average BMI, high rates of thinness among 1-19-year-olds, and elevated under-5 mortality. Authorities should concentrate on developing African and Asian countries to elevate life expectancy.